Baseten releases bison-1 and groundhog-1 decision models
Baseten launched bison-1 and groundhog-1, open-weight decision models built on Google’s Gemma 4 backbones. Both support text and image input, offer 128k-token context windows, and are available through Baseten Model APIs and with open weights on Hugging Face. Baseten says bison-1 ranked first among 19 models on Artificial Analysis Decision Index v2.1 with a score of 76 and recorded the index’s lowest Brier score.
The release adds open-weight, multimodal decision models with 128k context to Baseten’s API lineup, while its benchmark leadership claim gives the models a claimed advantage in accuracy and calibration.
Meet bison-1 and groundhog-1, frontier American decision models

Please rotate your device
To continue exploring this website, please rotate your device to a vertical orientation
AI models
Meet bison-1 and groundhog-1, frontier American decision models
Meet bison-1 and groundhog-1: open decision models on an American backbone (Gemma 4). bison-1 is the #1-ranked decision model on Artificial Analysis.
#Baseten#bison-1#groundhog-1#Gemma 4#Artificial Analysis#Baseten Model APIs#Hugging Face